Constrained Optimization of CMOS Analog Circuits via All-Inversion Region MOS Model

In this paper we furnish a novel approach of design automation and optimization of CMOS analog integrated circuits. This method offers an effective amalgamation between the principles of orthogonal-convex optimization and the all-inversion region MOS transistor model. We have observed that the constituting equations of the all-inversion region MOS model share great resemblance with the standard representation of orthogonal-convex functions. Therefore, these design equations, emerging from various device and circuit specifications can be modelled as the constraints of an orthogonal-convex optimization problem and can be evaluated automatically to ensure a globally optimal solution over a range of design scenarios. Additionally, there is a provision for design feasibility analysis with this semi-empirical approach.